eXist XML Database Overview Leif-Jran Olsson Introduction Current - - PowerPoint PPT Presentation

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

eXist XML Database

Leif-Jöran Olsson

Språkbanken, Department of Swedish Language, University of Gothenburg

2009-05-14

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Overview

◮ W3C Standards ◮ Current development foci ◮ Tuning ◮ Imdi use ◮ Some (other) uses ◮ Your topics of choice

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Introduction

◮ experience check ◮ W3C Standards

◮ xml ◮ xslt ◮ xquery ◮ xpath ◮ xmlschema

◮ web 2 ◮ modularisation, extensions

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Data model

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Examples: xpath et al

Axes and steps

nodes: node(), element(), attribute(), text(), comment(), processing-instruction() etc. Remember root-node(), document-element(), string-value document order context item/node primary axes: descendant, ancestor, preceding, following, self default axis: child abbreviated location steps: // (descendant-or-self::node()), / (child::node()), . (self::node()), .. (parent), @ (attribute), * child (elements), @* child attributes steps combined makes (location) paths

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Examples: xpath et al (2)

Effective boolean value (ebv) e g in predicate

$nodes[. = $sequence]

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Current development foci

◮ make the core lean i e moving out things from

the core, removing unnecessarily heavy dependencies

◮ pushing use of extensions, all types i e full text

searching with lucene for indexing, versioning, acl etc

◮ XQueryUrlRewrite (no need for cocoon) ◮ Statistically supported optimisations (maybe

even automatic index generation or at least suggestions)

◮ index support for aggregate functions (min,

max, order by, distinct-values etc.)

◮ collatable string-values ◮ physical partitioning ◮ Native xslt 2 processor ◮ Documentation (like functX) ◮ XQProc reference implementation in XQuery ◮ exquery, leveraging xquery function libraries

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Tuning

◮ http://exist-db.org/tuning.html

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

sign language corpus index (from IMDI)

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

lexemes with mother fish (from SALDO)

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Sentence count for corpus konkplus (from SBKHS)

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

What do you want?

Overview Introduction Current development foci Tuning Imdi use Some (other) uses Your topics of choice Final Comments

Final Comments

◮ use indexes ◮ Web services enable instant updates. ◮ However, it is important with some kind of version

control for the sake of reproducability.